Medium optimization for pyrroloquinoline quinone (PQQ) production by Methylobacillus sp. zju323 using response surface methodology and artificial neural network–genetic algorithm

Peilian Wei,Zhenjun Si,Yao Lu,Qingfei Yu,Lei Huang,Zhinan Xu
DOI: https://doi.org/10.1080/10826068.2017.1315596
2017-05-16
Abstract:Methylobacillus sp. zju323 was adopted to improve the biosynthesis of pyrroloquinoline quinone (PQQ) by systematic optimization of the fermentation medium. The Plackett-Burman design was implemented to screen for the key medium components for the PQQ production. CoCl<sub>2</sub> · 6H<sub>2</sub>O, ρ-amino benzoic acid, and MgSO<sub>4</sub> · 7H<sub>2</sub>O were found capable of enhancing the PQQ production most significantly. A five-level three-factor central composite design was used to investigate the direct and interactive effects of these variables. Both response surface methodology (RSM) and artificial neural network-genetic algorithm (ANN-GA) were used to predict the PQQ production and to optimize the medium composition. The results showed that the medium optimized by ANN-GA was better than that by RSM in maximizing PQQ production and the experimental PQQ concentration in the ANN-GA-optimized medium was improved by 44.3% compared with that in the unoptimized medium. Further study showed that this ANN-GA-optimized medium was also effective in improving PQQ production by fed-batch mode, reaching the highest PQQ accumulation of 232.0 mg/L, which was about 47.6% increase relative to that in the original medium. The present work provided an optimized medium and developed a fed-batch strategy which might be potentially applicable in industrial PQQ production.
biotechnology & applied microbiology,biochemical research methods,biochemistry & molecular biology
What problem does this paper attempt to address?